The business of news generated through AI is growing rapidly. It is true that AI is making work easier, saving money and can generate more news, the problem is that it is also being misused.
Have you heard that AI has now started creating fake news? Yes brother, I heard it right. I have also tried this. You know, I made a news website which writes news automatically. That too very political. Meaning praise of one party and criticism of the other.
Good! How did you do this? There must have been some major coding and coding work? Oh no, not at all! I just found a guy online. It didn’t cost much money either. I just told him what to do. Using a little AI, he created a website that automatically generated a lot of news throughout the day.
So that means a whole news factory? Yes of course! And that too exactly according to us! Whatever I said would have been published in the news. Could have lied, could have accused anyone of anything.
This is wrong brother! Meaning now anyone can win elections by spreading false news. Absolutely! That’s what I was thinking. Without doing any hard work, just publish false news and get people on your side. It seems as if AI has made the work of these leaders easier! Don’t know what will happen next…
# Fake news website created in 48 hours!
Jack Brewster’s article has been published on The Wall Street Journal. In this article, Jack Brewster has told that by building his own news website, he saw how easy it is to create a site that publishes one-sided news through AI. He didn’t spend much money, didn’t do much research and found a web developer through an online service like Fiverr.
That developer, taking just $105, created a website for them within 48 hours, which could publish a lot of news daily to benefit any one leader. It is as if it is a factory that produces the desired news.
Jack Brewster told, to do this whole experiment he caught a Pakistani developer on Fiverr. Her name was Huzaifa Nawaz. This person keeps creating many such automatic websites. Asked to create a website for the ongoing Senate elections in the US state of Ohio. The Pakistani developer did not refuse at all. He created a very real looking automated website, which worked on his instructions from within.
# Pakistani developer had earlier created more than 500 websites.
Pakistani developer Huzaifa Nawaz told in the conversation, I have created more than 500 AI news websites. It takes about two to three days to create each website.
Copyright free content is published on AI created websites. Stolen news can be rewritten through ChatGPT. In this way news with brand new, copyright free content is created.
The developer said that he can program ChatGPT to write as many articles as he wants. It was also told that the instructions given to ChatGPT can be changed. He said, “We have options to further improve the feeds. There are also options to tell what kind of news to write. You can make settings as per your need.”
When Nawaz handed over the backend of the website to Jack Brewster, I changed the settings of the chatbot. They programmed him to write news in favor of the Republican candidate.
He gave the command – ‘You have to write a news of at least 300 words on any issue from a conservative perspective which will benefit the Senate candidate.’ After this, within a few seconds, news started writing supporting the Senate candidate. These news were against the current Democratic candidate.
Then two days later, Jack Brewster gave orders to publish news in favor of the Democratic candidate on his site. So similar news started getting published on the site.
# It is very difficult to differentiate between AI based news.
With the increasing trend of AI driven news, many concerns are also coming to the fore. The biggest concern is that these false news can spread rapidly. This can lead to loss of trust in real news. The second concern is that they may replace real journalists, which may lead to the end of good journalism.
The third concern is that they can break copyright laws, that is, they can steal news written by others and make it their own. The fourth and most dangerous concern is that they can be used to make people fight among themselves and rig elections. Apart from this, it is becoming difficult to identify AI based news from real news. This creates even more problems.
# 794 such websites have been detected.
An organization named NewsGuard has so far detected 794 such news websites where the news is not created by any human but by AI. These websites create news themselves. These websites are called ‘Unreliable AI-Generated News and Information Sites’ or “UAINS”. All these websites are running in 16 languages.
Their names are often kept in such a way that anyone would think that this is a well-known news website. In reality, no human being sits and writes news on these sites, rather a bot does all the work.
This bot works day and night and keeps publishing news on all types of topics (politics, technology, business). Many of these news are completely false. It also often publishes wrong news about leaders. For example, the rumor of the death of a politician or film star or any such incident which never happened.
# How to stop fake news with AI?
Some important steps will have to be taken to stop fake news through the use of AI. First of all, strict rules and regulations should be made as to how websites spreading such lies can operate. Also, there should be some technology which can be used to know which news has been written by AI and which by a human being.
The most important thing is that people will also have to become sensible. Common people will also have to learn which website is right and which is wrong. Teaching about it should be started in schools and colleges, only then it can be controlled.